In a cannibalistic society, it’s not about survival – it’s all about recycling

A blog post highlighting the article by Chouvenc in Insectes Sociaux.

By Thomas Chouvenc

In this spooky time of the year, there are many examples we could draw from insect societies to give the heebie-jeebies to the non-entomophile. One to usually make it to the news cycle is, of course, the case of zombie-ants. It’s a classic dive into Z-culture while exploring the amazing biology of a host-manipulating fungal parasite. It’s hard to beat as click-bait because it’s just so good. Next in line would be parasitoid wasps, as their gruesome life cycle rips through the host’s organs from the inside while keeping it alive the entire time. Hollywood and manga artists have extensively dug into this concept to bring to life your favorite monster movies and books. Less known, but still a favorite for many of us: the Nicrophorus burying beetle. These Silphidae lay eggs on a decomposing carcass and display extensive parental behavior to their growing larvae as they chew through the putrid flesh of the roadkill. Gruesome, yet full of love: it’s the pinnacle of cute.

So, yes, Halloween could definitely use insects within the spectrum of sociality as core material for horrific displays in our front yards.  With all the classic anatomic fails of course, but this is another horror story. By the way, I was very disappointed when nothing came up from a “parasitoid wasp Halloween display” google search (someone should do something about this, please).

In my opinion, one example that has been either ignored or largely misrepresented in pop culture is insect cannibalistic societies. I will pass on the terrible cliché of old B movies on so-called “cannibal tribes”, but instead bring your attention to termites. Yes, the mostly-ignored or largely-misrepresented group of social insects (wink).  Termites have evolved along with the rise of angiosperms for ~150 Myr, eating a unique niche of untapped dry matter stored in tree trunks that is inaccessible by most plant-feeding animals. Eventually, termites evolved away from their Cryptocercus-like wood roach ancestor and reached the highest level of social organization. They were able to grow large colonies over time while optimizing their ecological success. However, while they enjoyed their pseudo-monopoly on woody material exploitation, their biology was constrained by a significant dietary restriction: wood is carbon-rich but notably nitrogen-poor.


Young and happy termite colony with a healthy king and queen, egg mass, brood, and workforce.

As a consequence, in basal termite taxa, colony metabolism is relatively slow. These taxa rely on their gut microbiota to do the hard work of cellulose degradation. Additionally, they take “forever” to grow, owing to limited nitrogen availability. This whole dilemma has been recognized as fundamentally important for the initial rise of eusociality and subsequent radiation of termites (Nalepa 1984). Termites could obtain all the carbon they needed to fuel the colony by feeding on ubiquitous wood. However, they had to optimize their entire metabolic engine toward nitrogen acquisition, and even more critical: nitrogen conservation. Termites have, therefore, perfected a recycling strategy toward nitrogen conservation over evolutionary time: cannibalism.


Not-so-happy termite colony (Coptotermes gestroi) about to run out of food.

To my knowledge, all termite species perform a colony-wide process of cannibalizing their sick, moribund, and dead individuals. This behavior has received extensive scrutiny by termite researchers around the world because it is so common and evident in termites that one can’t ignore it. Fun fact: when you study the survival of a group of termites in an experiment, you have to count the live termites, not the dead ones, simply because there is rarely a dead body lying around. It is usually cannibalized by nestmates very rapidly.  The concept of “refuse pile”, which can be rather crucial in ants, is mostly useless in termites, as everything goes through someone’s gut, eventually. Cannibalism in termite is, therefore, an essential part of the cycle of nutrients within the colony. It allows precious resources like nitrogen to be reused by the society. This behavior also has attained secondary functions such as sanitation and reducing the risk of diseases within the social group (Chouvenc et al. 2009).

The relationship between cannibalistic behavior and starvation in termites has also previously received some attention. It was suggested that cannibalism can have an essential role in helping a group of termites survive a period of starvation. As soldiers are nutritionally-dependent on workers, they would be eaten in priority. This cannibalism would alleviate their trophic burden to the group and recycle very precious resources toward the rest of the group. Also, the termite group would suppress the production of new soldiers so that only workers would remain, in the hope of restarting the metabolic engine of the group and reducing the trophic burden even further (Su and Lafage 1986). This strategy would, therefore, reduce the metabolic footprint of a colony with finite resources, and aid in more prolonged colony survival.  Soylent green, allons-y!

In a study published online this month in Insectes Sociaux (Chouvenc 2019), I revisited this concept, but this time, I took the “whole colony” approach. Previously, researchers had subjected a small group of foraging termites collected from the field to starvation. To improve upon this method, I subjected 10 whole young termite colonies (Coptotermes gestroi) to starvation. I wanted to take the investigation to a more biologically-relevant scale, with ~3,000 termites per colony, a healthy brood, overlapping generations of workers and soldiers, and of course, the king and queen. The effects of starvation on larger termite groups, the brood, or the primary reproductive were never before investigated.

Recently, I have started revisiting some old questions but scaling things up to the colony level. It can be challenging not to be disappointed in seeing a somewhat different outcome than the one initially expected. To make a long story short: termites are terrible in their survival strategy during starvation events because, in fact, they don’t really have a strategy. Unlike honeybees that store months-worth of honey, termites have a carpe-diem approach to food safety, as they have no internal reserves. Instead, subterranean termites such as C. gestroi will relentlessly (and most of the time successfully) forage for new food sites to prevent food shortage in the first place. But if starvation of the termite colony actually occurs, then the colony just doesn’t have much of a survival strategy. Give it about 30 days.

As the metabolism of the colony is progressively running out of fuel, nutritionally stressed individuals start accumulating in the colony. Unfortunately for termite larvae and workers, who are hemimetabolous insects stuck in a permanent juvenile molting cycle, the time to molt eventually comes, and the younger the instar, the faster the molting cycle. Have you ever tried molting while completely starving?  I would not advise. In my study, this resulted in a failed attempt to molt, death, and subsequent cannibalism from nestmates.

Therefore, the brood and young workers were the first ones to be eaten, not as survival rations for the group, but because they were the first ones to die during their molting process. Then soldiers started running out of juice faster than some of the older workers (who have yet to attempt molting), as they are fed secondhand from workers. Indeed, the remaining workers had to maintain their own metabolism and had nothing left to share with the soldiers, resulting in moribund soldiers, and inherent subsequent cannibalism. In addition, it’s not that the starving colony suppresses the production of new soldiers. It is just that presoldiers that are starving also failed to molt into functional soldiers, and therefore died and were eaten, resulting in an apparent absence of soldier replacement.

Finally, toward the end, a handful of workers remained, with the king and queen being the last ones to die, inevitably.  So, if termites actually have a survival strategy during starvation, it is this: “keep the king and queen alive as long as you can”. Any remaining available energetic resources were eventually funneled to them.

In the end, I found that, contrary to a previous perception, termites don’t reduce the trophic burden of the colony by cannibalizing the dependent castes. Mortality was not cannibalism-driven; instead, cannibalism was mortality-driven.  Termites just do what they always do: if a dead or moribund individual shows up in the group, it is cannibalized to recycle the nitrogen. This is an inherent behavior that was reinforced over millions of years of a nitrogen-deficient diet. The fact is, cannibalizing an energy-depleted individual may not provide much energy to the group, as you can’t recycle ATP that does not exist.

The excessive mortality resulting from starvation triggered a massive cannibalism wave, which ended up with an accumulation of old workers with staggering levels of uric acid building up in their fat body. Such an observation is typical of termite colony collapse, as there is a sudden excess of available nitrogen through cannibalism, something termite metabolism never evolved toward.  By 20 days after the starvation was in effect, the colony started shutting down, and cannibalism was no longer observed. Dead bodies began accumulating in the colony and were not taken care of by the surviving workers, which were focusing only on the king and queen. In the end, the king and queen eventually starved to death too. So much for an efficient survival strategy in termites.

In this study, despite the amount of work it took to investigate a question at the colony-level and provide detailed observation, it was, in my opinion, absolutely worth it. Experimenting on various aspects of termite biology in small groups in a Petri dish can provide valuable initial information. However, scaling it up to a colony-wide observation can drastically change our understanding of a social group.  For example, if you have a termite infestation in your house, you can now actually say “my house is being attacked by a horde of cannibals.” It would be almost accurate. Happy Halloween.


Although this paper discussed how cannibalism can go wrong in a starving termite colony, this picture reveals a case of cannibalism gone (very) weird – if one can make any sense out of it. It depicts a worker in the process of cannibalizing a nestmate, but his head is stuck inside the terminal abdominal segments of the dead one. Photo courtesy from David Mora Del Pozo.

Thomas Chouvenc is an Assist. Professor in urban entomology at the University of Florida, UF/IFAS.

Twitter: @ChouvencL


Chouvenc T (2019) Limited survival strategy in starving subterranean termite colonies. Insectes Sociaux.

Chouvenc T, Su NY, Robert A (2009) Inhibition of Metarhizium anisopliae in the alimentary tract of the eastern subterranean termite Reticulitermes flavipes. J Invertebr Pathol 101: 130-136

Nalepa CA (1994) Nourishment and the origin of termite eusociality. Nourishment and Evolution in Insect Societies (ed. by: J.H. Hunt & C.A. Nalepa), pp. 57-104. Westview Press, Boulder, Colorado

Su NY, La Fage JP (1986) Effects of starvation on survival and maintenance of soldier proportion in laboratory groups of the Formosan subterranean termite, Coptotermes formosanus (Isoptera: Rhinotermitidae). Ann Entomol Soc Am 79: 312-316

Leave a Reply

Fill in your details below or click an icon to log in: Logo

You are commenting using your account. Log Out /  Change )

Facebook photo

You are commenting using your Facebook account. Log Out /  Change )

Connecting to %s